AI-infused patients for enhanced clinical simulation

Artificial intelligence (AI) refers to the simulation of human intelligence in computers, allowing them to perform tasks that usually require human cognitive abilities, such as decision-making, problem-solving, and learning (Pesapane et al., 2020). AI is transforming medical education by enhancing how students learn. AI’s supporting function in medical, nursing, and allied health programs is rapidly expanding worldwide, as institutions recognize its potential to create more engaging, adaptive learning experiences that address individual learning needs. By improving both curriculum design and learner assessment, AI is helping shape a more effective, accessible, and personalized approach to training the next generation of healthcare professionals. Advantages of AI Implementation in Medical Education (Mahdi Zarei et al.): Enhanced Curriculum Design and Evaluation. AI enables improved curriculum development and evaluation, facilitating the implementation of innovative educational methods in clinical settings. Integration with Virtual Reality (VR). The use of AI in medical education is often accompanied by the expansion of virtual reality, providing immersive learning environments. Efficient Assessment Processes. AI can streamline formative and summative assessments, reducing time and costs. Personalized Feedback and Guidance. By offering individualized feedback, AI enriches the learning experience, tailoring guidance to each student’s needs. Stress-Free Simulation Environments. Clinical simulators powered by AI provide stress-free environments where students can practice medical reasoning and learn from mistakes. Increased Diagnostic Proficiency. AI-based technologies can enhance students’ diagnostic skills by up to 22%. Enhanced Transparency and Comparability. The use of AI in medical education increases transparency and enables meaningful comparisons of educational effectiveness across different countries and institutions. Expanded Access to Education. AI facilitates distance learning, making medical education more accessible in resource-limited and remote areas. Continuous Learning Opportunities. AI offers students the chance to learn continuously, 24/7, enabling flexible study schedules. Support for Active Learning Strategies. AI can facilitate active learning approaches, such as problem-based, case-based, small-group, and large-group learning. Improved Identification of Student Needs. AI improves the quality of education by identifying and addressing individual learning needs. Cost-Effective Quality Enhancement. AI-based technologies support curriculum reform and the implementation of innovative teaching methods in a cost-effective manner. As medical education increasingly integrates advanced technologies, the role of AI in clinical simulation has emerged as a critical tool for preparing healthcare providers. AI-infused patients offer an interactive, risk-free environment where learners can practice clinical skills, receive real-time feedback, and engage in scenarios that adapt dynamically based on their decisions. This presentation explores the development, application, and impact of AI-infused patients in healthcare training, positioning them as a transformative resource in clinical education. The integration of AI in clinical skills education encompasses various elements, including AI and Standardized Patients (SPs), AI and Virtual Patients, AI and Virtual Instructors, and AI and Intelligent Instructional Tutoring Systems, as discussed by Hamilton A. in the article "Artificial Intelligence and Healthcare Simulation: The Shifting Landscape of Medical Education". One significant aspect is text creation, where AI facilitates the development of personas and chatbots, enabling the generation of realistic dialogue between patients and healthcare professionals (Marc Lazarovici. SESAM 2024 Workshop: "The Art and Craft of ChatGPT"). This includes crafting patient responses and generating questions that doctors might ask during consultations. Furthermore, AI aids in the creation of consultation guides for Objective Structured Clinical Examinations (OSCE). In addition to text-based tasks, AI plays a crucial role in simulation scenario creation (example of a Simulation Scenario Creation Tool by Frederico Lorenzo Barra). It assists educators in drafting scenario stems, formulating learning objectives, recommending the necessary equipment and resources, structuring debriefing points, and providing relevant topical references that enrich the learning context. Assessment and debriefing processes are also enhanced through AI capabilities, enabling a more comprehensive evaluation of student performance and learning outcomes. Moreover, AI contributes to the visual aspects of education, facilitating the creation of images and video avatars that can simulate real-life scenarios, further immersing learners in their training experiences. This multifaceted approach demonstrates the transformative potential of AI in advancing healthcare education and simulation practices. The review process conducted by human experts on simulation scenario creation, as detailed in the study by Rodgers et al. (2023), highlighted several strengths and areas for improvement in the development of scenarios by AI. Among the strengths, the scenarios were praised for their specific objectives, which provided clear direction for participants. The debriefing plans were noted for being clear, thorough, and well-defined, ensuring that participants could effectively reflect on their experiences. Additionally, the scenarios were aligned with simulation standards and successfully created real-world contexts, enhancing their relevance. The definition of participant roles was also a strong point, as it included potential challenges that participants might face. Furthermore, the inclusion of medication dosing guidelines and pediatric cases was appropriate for the intended audience, contributing to the scenarios' educational value. However, the review identified several areas for improvement. One significant concern was the lack of performance-branched patient state progression, which could limit the adaptability of scenarios to different participant responses. Some objectives were found to be not time-bound, which may affect the pacing of the simulation. Additionally, inaccuracies in medication dosing were noted, along with a failure to reference current treatment guidelines. Specific treatment algorithms were not included, and not all key equipment and supplies were accounted for, potentially hindering the realism of the scenarios. Furthermore, it was observed that objectives for adult cases might be too advanced for the audience, and some references provided were either inaccurate or confabulated. Overall, the review by Rodgers et al. underscores the importance of continuous evaluation and refinement in simulation scenario creation to enhance educational effectiveness in healthcare training. According to research by Leticia De Matte and coauthors titled Are Artificial Intelligence Virtual Simulated Patients (AI-VSP) a Valid Teaching Modality for Health Professional Students? published in Clinical Simulation in Nursing (Volume 92, 2024), a majority of students responded positively to various Likert scale items related to their interactions with AI-based virtual simulated patients (AI-VSPs). Students generally agreed with statements such as, “My interaction with the AI-patient felt realistic (e.g., environment, avatar, voices),” “I gave this virtual patient my full attention and developed a diagnosis and plan of care for the chief complaint just as thoroughly as I would have done with a human patient in clinic,” “I feel virtual patient experiences could help me improve my diagnostic ability,”  “This scenario helped me with history-taking and developing a differential diagnosis shortlist”  and “I would recommend AI-patient to other healthcare learners interested in practicing history-taking.” The majority agreed with these statements, indicating a generally favorable perception of AI-VSPs as realistic, engaging, and beneficial tools for practicing clinical skills, compared to those who were neutral or disagreed. At the same time, cognitive branching simulation holds significant value in clinical training by allowing learners to navigate complex decision-making pathways, mirroring real-life clinical situations. This not only promotes critical thinking and problem-solving skills but also facilitates individualized learning experiences tailored to each learner's unique decision-making process. We aim to bridge the concepts of cognitive branching and AI in medical education. We focus on developing cognitive branching simulations that utilize adaptive strategies for evolving events within clinical scenarios, fundamentally connected to AI logic. This approach serves as a foundation for AI-driven medical simulations, enhancing the learning experience for healthcare professionals. For cognitive branching simulation in the simulation training platform "ClinCaseQuest" "A Mathematical Model Based on Stratifying the Severity of Medical Errors for Building Scenarios for Clinical Cases With Branching” (Lopina N. 2024, Cureus) and “A Staged Defragmented Simultaneous Debriefing Model As Integrated Micro-Debriefing Components Inside Online Simulation for Competencies Formation” (Lopina N. 2024, Cureus) were developed to construct branching clinical scenarios. Together, these works lay the groundwork for advancing cognitive branching simulations, highlighting their potential to enhance the effectiveness and adaptability of medical training through the potential integration of AI technologies. The report presented examples of using ChatGPT for image creation in branching simulations, utilizing Consistent Character GPT, which helps generate consistent images of the patient. Prompt example: “Create a patient in the Hospital Room. Description: A highly detailed photorealistic scene of a hospital room with a 53-year-old female patient named Mary. Mary has short brown hair and is overweight, weighing 94 kg and standing 174 cm tall. She is lying in a hospital bed, wearing light-blue patient clothes. A nasal oxygen cannula is in place for oxygen therapy. The background shows typical hospital equipment, including a monitor and IV stand. The room has soft, natural lighting coming through a window, adding a sense of calm. Setting: A hospital room with white walls, medical equipment, and a window with soft natural light. Style: Photorealistic. Image Ratio: 16:9“ Other examples include the clinical case simulator “An Interrupted Symphony”, which focuses on managing patients with symptomatic bradycardia. Another example is the clinical case simulator “Mysteries of the Heart”, dedicated to Takotsubo syndrome. In the evolution of clinical education, AI-infused patients represent a pivotal advancement, merging artificial intelligence with simulation-based learning to create virtual patients capable of nuanced interactions, responsive decision-making, and dynamically branching scenarios. These AI-driven entities simulate human behaviors and reactions, allowing learners to engage in complex, realistic interactions that help bridge the gap between theoretical knowledge and real clinical practice. Unlike traditional static simulators or mannequin-based simulations, AI-infused patients can communicate, and respond to clinical inquiries, enhancing the authenticity of the learning experience. AI-infused patients are built on several key technological foundations, each contributing to their realism and educational value. Here are the primary characteristics: Natural Language Processing (NLP): Role in Communication. NLP allows AI-infused patients to understand and respond to spoken or written language inputs. This feature enables learners to ask questions, give instructions, and receive verbal responses from AI-patients, simulating real clinical interactions. Real-Time Dialogue Generation. AI-patients use NLP to create natural, fluid dialogues, allowing for questions, clarifications, and follow-ups. This feature gives students a chance to practice patient interviews, collect medical histories, and probe for more information, enhancing clinical reasoning skills. Personalization and Contextual Responses. Through advanced NLP, AI patients can recognize context-specific cues and adapt their responses, providing a level of interaction that feels tailored to each student’s questions and decisions. Data Analytics and Feedback Mechanisms Performance Analysis. AI-patients are equipped with data analytics capabilities, allowing educators to track and assess student performance on various metrics. These metrics can include response times, accuracy in diagnosis, scores, and adherence to clinical protocols. Individualized Feedback. At the end of each simulation, learners receive personalized feedback based on their interactions. This feedback can highlight strengths and areas for improvement, reinforcing the learning objectives of the exercise. Objective Assessment for Skills Development. By providing data-driven assessments, AI-infused patients enable a more objective measurement of skills, making it easier to identify areas for curriculum improvement and individualized student support. While traditional simulators - such as high-fidelity mannequins and standardized patients - have long been used in healthcare education, AI-infused patients offer a level of versatility, realism, and adaptability that marks a new era in medical training. Dynamic Response Capabilities. Traditional mannequins may provide programmed responses or reactions, but they lack the adaptive intelligence of AI-infused patients. Unlike pre-scripted simulations, AI-patient scenarios evolve in real time, allowing students to explore multiple treatment paths and observe varied outcomes. Scalability and Accessibility. AI-driven simulations are easily scalable and accessible remotely, unlike traditional simulators requiring physical presence. This remote capability is invaluable for distance learning programs, enabling consistent training quality across geographies. Resource Efficiency.  AI-infused patients streamline the educational process by reducing the need for human actors and other costly resources in simulation labs. They offer an economical solution that can be widely implemented, especially beneficial in regions with limited resources. AI-infused patients represent a transformative shift in healthcare education, allowing medical students and professionals to engage in safe, immersive learning experiences that significantly enhance clinical skills. These advanced simulation tools, which utilize artificial intelligence to emulate real-life patient interactions, offer substantial educational benefits that go beyond traditional methods. By promoting hands-on learning, providing adaptive and individualized feedback, and facilitating critical thinking, AI-infused patients prepare healthcare providers for the challenges they will face in real clinical settings. Combining branching scenario generation with artificial intelligence technologies represents a promising direction for enhancing medical education. This approach leverages AI’s capabilities to create dynamic, adaptive learning experiences that can better simulate real-life clinical situations. In branching scenarios, learners are presented with decision points that lead to various outcomes based on their choices. By integrating AI, these scenarios can be made more sophisticated and responsive. For instance, AI can analyze learners' actions and adapt the scenarios in real time, providing personalized feedback and altering the patient’s condition or the complexity of the situation based on the learner’s decisions. This not only helps in assessing clinical reasoning and decision-making skills but also encourages deeper engagement by making the learning experience more immersive and relevant. Moreover, AI technologies can enhance the generation of realistic dialogue and interactions within these scenarios. They can create diverse patient personas with unique medical histories, presenting challenges that reflect the variability of real-world cases. This ability to tailor scenarios based on learner performance and preferences allows for more targeted training, ultimately improving clinical competency. Furthermore, AI can assist educators in the design and evaluation of these scenarios. By analyzing data from previous simulations, AI can identify patterns and suggest improvements, ensuring that scenarios remain current and aligned with best practices in medical education. Overall, the integration of branching scenario generation with AI technologies holds the potential to revolutionize medical training, offering adaptable, engaging, and effective educational tools that prepare learners for the complexities of clinical practice. Conclusion of Advantages of AI-Infused Patients AI-infused patients offer a wide range of advantages for healthcare training, from improving clinical competency and diagnostic accuracy to fostering empathy and communication skills. By providing realistic, adaptable learning experiences, AI-infused simulations address the educational needs of both learners and institutions, supporting a future in which healthcare providers are better prepared, more compassionate, and more competent in delivering patient care. As healthcare education continues to evolve, AI-infused patients are likely to play a central role in shaping the next generation of skilled professionals. References: Pesapane F, Tantrige P, Patella F, Biondetti P, Nicosia L, Ianniello A, Rossi UG, Carrafiello G, Ierardi AM. Myths and facts about artificial intelligence: why machine- and deep-learning will not replace interventional radiologists. Med Oncol. 2020 Apr 3;37(5):40. doi: 10.1007/s12032-020-01368-8. PMID: 32246300. Mahdi Zarei, Hamid Eftekhari Mamaghani, Amin Abbasi, Mohammad-Salar Hosseini. Application of artificial intelligence in medical education: A review of benefits, challenges, and solutions, Medicina Clínica Práctica, Volume 7, Issue 2, 2024,100422, ISSN 2603-9249, https://doi.org/10.1016/j.mcpsp.2023.100422. Hamilton A. Artificial Intelligence and Healthcare Simulation: The Shifting Landscape of Medical Education. Cureus. 2024 May 6;16(5):e59747. doi: 10.7759/cureus.59747. PMID: 38840993; PMCID: PMC11152357. Marc Lazarovici. SESAM 2024 Workshop: "The Art and Craft of ChatGPT" Supplementary Material . Simulation Scenario Creation Tool by Frederico Lorenzo Barra.    Rodgers DL, Needler M, Robinson A, Barnes R, Brosche T, Hernandez J, Poore J, VandeKoppel P, Ahmed R. Artificial Intelligence and the Simulationists. Simul Healthc. 2023 Dec 1;18(6):395-399. doi: 10.1097/SIH.0000000000000747. Epub 2023 Sep 20. PMID: 37747487. Leticia De Mattei, Marcelino Q. Morato, Vineet Sidhu, Nodana Gautam, Camila T. Mendonca, Albert Tsai, Marjorie Hammer, Lynda Creighton-Wong, Amin Azzam, Are Artificial Intelligence Virtual Simulated Patients (AI-VSP) a Valid Teaching Modality for Health Professional Students? Clinical Simulation in Nursing, Volume 92, 2024, 101536, ISSN 1876-1399, https://doi.org/10.1016/j.ecns.2024.101536. Lopina N. A Mathematical Model Based on Stratifying the Severity of Medical Errors for Building Scenarios for Clinical Cases With Branching. Cureus. 2024 Apr 11;16(4):e58089. doi: 10.7759/cureus.58089. PMID: 38738126; PMCID: PMC11088723. Lopina N. A Staged Defragmented Simultaneous Debriefing Model As Integrated Micro-debriefing Components Inside Online Simulation for Competencies Formation. Cureus. 2024 Mar 12;16(3):e56000. doi: 10.7759/cureus.56000. PMID: 38606236; PMCID: PMC11007450. Example AI generated design for branching scenario clinical case simulator “An interrupted symphony” (Managing Patients with Symptomatic Bradycardia) https://clincasequest.hospital/course/interrupted-symphony/ Example AI generated design for branching scenario clinical case simulator “Mysteries of the heart” (Managing Patients with Takotsubo syndrome) https://clincasequest.hospital/course/mysteries-of-the-heart/